Bioorganic Chemistry 37 (2009) 167–172

Contents lists available at ScienceDirect

Bioorganic Chemistry

journal homepage: www.elsevier .com/locate /bioorg

Kinetic evidence for an anion binding pocket in the active siteof nitronate monooxygenase q

Kevin Francis a, Giovanni Gadda a,b,c,*

a Department of Chemistry, Georgia State University, Atlanta, GA 30302-4098, United Statesb Department of Biology, Georgia State University, Atlanta, GA 30302-4098, United Statesc The Center for Biotechnology and Drug Design, Georgia State University, Atlanta, GA 30302-4098, United States

a r t i c l e i n f o

Article history:Received 23 June 2009Available online 25 July 2009

Keywords:2-Nitropropane dioxygenaseNitronate monooxygenaseAnion inhibitionAlkyl nitronatesCompetitive inhibitionHansenula mrakiiNeurospora crassaSubstrate specificity

0045-2068/$ - see front matter � 2009 Elsevier Inc. Adoi:10.1016/j.bioorg.2009.07.005

Abbreviations: NMO, nitronate monooxygenase; Fq This work was supported in part by Grant PRF #4

Chemical Society (to G.G.) a Molecular Basis of DiseasePendergrast Fellowship from Georgia State University

* Corresponding author. Address: Georgia State Ument, P.O. Box 4098, Atlanta, GA 30302-4098, Unitedfax: +1 404 413 5551.

E-mail address: ggadda@gsu.edu (G. Gadda).

a b s t r a c t

A series of monovalent, inorganic anions and aliphatic aldehydes were tested as inhibitors for Hansenulamrakii and Neurospora crassa nitronate monooxygenase, formerly known as 2-nitropropane dioxygenase,to investigate the structural features that contribute to the binding of the anionic nitronate substrates tothe enzymes. A linear correlation between the volumes of the inorganic anions and their effectiveness ascompetitive inhibitors of the enzymes was observed in a plot of pKis versus the ionic volume of the anionwith slopes of 0.041 ± 0.001 mM/Å3 and 0.027 ± 0.001 mM/Å3 for the H. mrakii and N. crassa enzymes,respectively. Aliphatic aldehydes were weak competitive inhibitors of the enzymes, with inhibition con-stants that are independent of their alkyl chain lengths. The reductive half reactions of H. mrakii nitronatemonooxygenase with primary nitronates containing two to four carbon atoms all showed apparent Kd

values of �5 mM. These results are consistent with the presence of an anion binding pocket in the activesite of nitronate monooxygenase that interacts with the nitro group of the substrate, and suggest a min-imal contribution of the hydrocarbon chain of the nitronates to the binding of the ligands to the enzyme.

� 2009 Elsevier Inc. All rights reserved.

1. Introduction

Nitronate monooxygenase (E.C. 1.13.11.32; NMO), formerlyknown as 2-nitropropane dioxygenase [1] is a flavin mononucleo-tide-dependent (FMN) enzyme that catalyzes the oxidative denitri-fication of alkyl nitronates to their corresponding aldehyde andketo compounds and nitrite [2,3]. The most extensively character-ized NMOs studied to date are those from Neurospora crassa [2,4–6]and Hansenula mrakii [3,7], although the X-ray crystallographicstructure of the enzyme from Pseudomonas aeruginosa has alsobeen reported [8]. Detailed mechanistic studies have been carriedout only for N. crassa NMO where it was shown that a transient an-ionic flavosemiquinone is formed during oxidative catalytic turn-over of the enzyme through a single electron transfer reactionbetween an enzyme-bound nitronate and the flavin cofactor(Scheme 1) [2,4]. The formation of an anionic flavosemiquinoneintermediate during oxidative catalysis is a characteristic feature

ll rights reserved.

MN, flavin mononucleotide.

3763-AC4 from the AmericanFellowship and a Ambrose H.(K.F.).

niversity, Chemistry Depart-States. Tel.: +1 404 413 5537;

of NMO that distinguishes it from the well characterized nitroal-kane oxidase [9], which catalyzes a similar oxidation reactionthrough a different mechanism that involves the formation of acovalent flavin N(5)-adduct [10–13].

Recombinant NMO from H. mrakii was recently cloned and ex-pressed in Escherichia coli cells and the resulting purified enzymewas characterized in its biochemical and kinetic properties [3].The enzyme is similar to that from N. crassa in that it contains asingle non-covalently bound FMN per monomer of enzyme and isdevoid of metal cofactors [2,3]. Moreover, an anionic flavosemiqui-none was observed upon anaerobic mixing of H. mrakii NMO withalkyl nitronates with chain lengths ranging from two to six carbonatoms [3]. Neither hydrogen peroxide nor superoxide is releasedduring turnover of the enzyme with primary alkyl nitronates asevident from the absence of superoxide dismutase or catalaseeffects on the rates of oxygen consumption in activity assays ofH. mrakii NMO [3], a result that was also obtained with the N. cras-sa enzyme [2].

Both H. mrakii and N. crassa NMOs are able to effectively oxidizea number of alkyl nitronates into their corresponding carbonylcompounds and nitrite. Such an enzymatic oxidation is of consid-erable interest given that many alkyl nitronates are known to betoxic or mutagenic [14–17]. Ingestion of propyl-2-nitronate, forexample, has been demonstrated to result in the formation of8-aminodeoxyguanosine and 8-oxodeoxyguanosine through a

C HH3C

NOO

N

N

NH

N

O

O

R

H3C

H3C

C HH3C

NOO

N

N

NH

N

O

O

R

H3C

H3C

C HH3C

NOO

N

N

NH

N

O

O

R

H3C

H3C

C HH3C

NOO

N

N

NH

N

O

O

R

H3C

H3C

Scheme 1. The one-electron oxidation of ethylnitronate catalyzed by nitronatemonooxygenase.

168 K. Francis, G. Gadda / Bioorganic Chemistry 37 (2009) 167–172

phenol sulfotransferase-mediated metabolic pathway in both hu-man and rat cell lines [18,19]. Despite their toxicity, alkyl nitro-nates are widely used in chemical industry because they providea quick and efficient route for the synthesis of a wide range of com-mercially useful compounds [20–22]. An investigation of the sub-strate specificity of NMO can therefore provide the basis for useof the enzyme in bioremediation applications to detoxify wastegenerated from industrial uses of alkyl nitronates.

In the current study, the contributions of both the nitro andhydrocarbon moieties of the alkyl nitronate substrate of NMO forbinding and specificity were investigated through inhibition andrapid kinetics studies. The results are consistent with the presenceof an anion binding pocket in both the H. mrakii and N. crassa en-zymes, which interacts with the nitro group of the substrate. Theseinteractions are key determinants for binding and recognition ofthe substrate by the enzyme, rather than the hydrophobic interac-tions that could occur between the enzyme and the alkyl moiety ofthe substrate, as in the case of nitroalkane oxidase [23,24].

2. Materials and methods

2.1. Materials

NMOs from H. mrakii and N. crassa were obtained through theexpression and purification protocols described previously [3,5].Nitroethane, 1-nitropropane and 1-nitrobutane were from Sig-ma–Aldrich (St. Louis, MO). All other reagents were of the highestpurity commercially available.

2.2. Steady state kinetics

Enzymatic activity was measured in 50 mM potassium phos-phate at pH 7.4 and 30 �C with the method of initial rates [25] bymonitoring the rate of oxygen consumption with a computer inter-faced Oxy-32 oxygen monitoring system (Hansatech InstrumentLtd.). Enzyme concentrations were expressed per bound FMN con-tent using experimentally determined values of 13,100 M�1 cm�1

(e446 nm) for the H. mrakii enzyme [3] and of 11,850 M�1 cm�1

(e444 nm) for the N. crassa enzyme [2]. The final concentration of en-zyme used in each assay was between 25 and 65 nM, whereas sub-strate concentrations ranged from 0.5 to 20 mM. The nitronateform of the substrate was prepared in 100% ethanol by incubatingthe corresponding nitroalkane with 1.2 M excess of potassiumhydroxide for at least 24 h at room temperature. Since the sec-ond-order rate constant for the protonation of ethylnitronate is15 M�1 s�1 [26], enzymatic activity assays were initiated by theaddition of substrate to the reaction mixture to ensure that a neg-ligible amount of the neutral form of the nitronate is formed duringthe time required to determine initial rates of reaction (typically�30 s).

2.3. Pre-steady state kinetics

The pre-steady state kinetic parameters of H. mrakii NMO weredetermined in 50 mM potassium phosphate at pH 7.4 and 30 �Cusing a TgK Scientific SF-61 stopped-flow spectrophotometer.Rates of flavin reduction were measured by monitoring the in-crease in absorbance at 372 nm that results from anaerobic mixingof the enzyme with substrate as previously described for N. crassaNMO [4]. Nitronate solutions (100 mM) were prepared in water byincubating the nitroalkane in a 1.2 M excess of potassium hydrox-ide for at least 24 h and were diluted in water prior to use. The finalenzyme concentration in each assay was between 9 and 20 lM,whereas the substrate concentrations used ranged from 0.1 to50 mM, thereby ensuring that the enzymatic reaction followspseudo first-order kinetics.

2.4. UV–visible absorbance spectra of NMO in the presence of sodiumnitrite

Changes in the UV–visible absorbance spectrum of NMO uponaddition of sodium nitrite were monitored using an AgilentTechnologies diode-array spectrophotometer Model HP 8453,thermostated at 15 �C. Spectra of the H. mrakii and N. crassa en-zymes (at concentrations of �80 lM) were recorded before andafter addition of 1 mM sodium nitrite. Difference spectra werethen constructed by subtracting the final absorbance spectrumof the enzyme in the presence of sodium nitrite from that ofthe free enzyme.

2.5. Data analysis

Steady state kinetic data were fit with either Enzfitter (Biosoft,Cambridge, UK) or KaleidaGraph software (Synergy Software,Reading, PA). Stopped-flow traces monitoring the reductive halfreaction of H. mrakii NMO were fit with Eq. (1), which describesa single exponential process where kobs is the observed first-orderrate for the increase in absorbance at 372 nm, At is the absor-bance at time t, and A is the final absorbance. Pre-steady statekinetic parameters were determined using Eq. (2), where kobs isthe observed rate of flavin reduction, kred is the limiting rate con-stant for flavin reduction at saturating substrate concentrations,and Kd is the apparent dissociation constant of the substrate (S).Inhibition data were fit with Eq. (3), which describes a competi-tive inhibition pattern where Kis is the dissociation constant forthe inhibitor (I).

Atotal ¼ Ate�kobst þ A ð1Þ

kobs ¼kredS

Kd þ Sð2Þ

vo

e¼ kcatS

Km½1þ ð IKisÞ� þ S

ð3Þ

3. Results

3.1. Nitrite inhibition of NMO with respect to ethylnitronate assubstrate

In order to establish if the nitro group of the akyl nitronate sub-strates of H. mrakii and N. crassa NMO contributes to binding andspecificity, nitrite was used as a mimic of the substrate to establishwhether it inhibits the enzymes in 50 mM potassium phosphatepH 7.4 and 30 �C. As shown in Fig. 1A for the case of the H. mrakiienzyme, sodium nitrite behaved as a competitive inhibitor with re-spect to ethylnitronate as substrate for both enzymes as indicated

Fig. 1. Inhibition of H. mrakii NMO by sodium nitrite with respect to ethylnitronateas substrate. Panel A: Inhibition of H. mrakii NMO. Enzymatic activity was measuredat varying concentrations of ethylnitronate in the presence of 0 mM (�); 5 mM (o);10 mM (j); 20 mM (h) and 50 mM (N) sodium nitrite in 50 mM potassiumphosphate at pH 7.4 and 30 �C. The lines are from fits of the data to Eq. (3). Panel B:UV–visible absorbance spectral changes induced upon addition of 1 mM sodiumnitrite to H. mrakii (top) of N. crassa (bottom) NMO in 50 mM potassium phosphatepH 7.4 at 15 �C. Difference spectra were constructed by subtracting the finalabsorbance spectrum of the enzyme in the presence of nitrite from that of the freeenzyme.

Table 1Inorganic anion inhibition of NMO with respect to ethylnitronate as substratea.

Inhibitor Ionic volume (Å3)b H. mrakii, Kis

(mM)cN. crassa, Kis

(mM)d

Sodium nitrite 55 1.7 ± 0.1 10 ± 1Potassium fluoride 25 77 ± 2 125 ± 7Potassium chloride 47 11.3 ± 0.3 30 ± 2Potassium bromide 56 4.0 ± 0.3 20 ± 1Sodium nitrate 64 1.9 ± 0.1 12 ± 1Potassium iodide 72 0.95 ± 0.06 6.0 ± 0.4

a Enzymatic activity was measured at varying concentrations of ethylnitronateand several fixed concentrations of inhibitor in 50 mM potassium phosphate pH 7.4at 30 �C. Data were fit with Eq. (3).

b From [27].c The steady state kinetic parameters had the following average values:

kcat = 180 ± 15 s�1; Km = 3.6 ± 0.2 mM and kcat/Km = 48,500 ± 6500 M�1 s�1.d The steady state kinetic parameters had the following average values:

kcat = 68 ± 1 s�1; Km = 1.6 ± 0.1 mM and kcat/Km = 42,700 ± 3300 M�1 s�1.

K. Francis, G. Gadda / Bioorganic Chemistry 37 (2009) 167–172 169

by the pattern of lines that intersect on the y-axis in double reci-procal plot of the initial rate of oxygen consumption versus ethyl-nitronate concentration at different fixed concentrations ofinhibitor1. The corresponding inhibition constants (Kis) were1.7 ± 0.1 mM for the H. mrakii enzyme and 9.8 ± 0.6 mM for theN. crassa enzyme. The effect of the nitrite on the UV–visible absor-bance spectra of the FMN cofactor of the H. mrakii and N. crassaNMOs was also determined. As shown in Fig. 1B, binding of sodiumnitrite to both enzymes induced spectral changes in the FMN cofac-tor bound at the active sites of the enzymes as seen from theincrease in the absorbance intensities at 333, 414, 436 and 465 nmalong with the concomitant decrease in absorbance intensities at381 and 485 nm. All taken together, the inhibition and spectroscopicstudies are consistent with binding of nitrite occurring at the activesite of the enzyme.

3.2. Inorganic anion inhibition of NMO with respect to ethylnitronateas substrate

A series of monovalent, inorganic anions were tested as inhibi-tors for NMO in 50 mM potassium phosphate pH 7.4 and 30 �C todetermine if the enzyme contains an anion binding site for recog-nition of the nitro group of the substrate. As shown in Table 1, inor-ganic anions ranging in size from 25 to 64 Å3 were competitiveinhibitors for the H. mrakii enzyme, with inhibition constants rang-

1 All of the inhibition data reported in this study were also fit with equationsdescribing noncompetitive and uncompetitive inhibition patterns. Since the datawere all best fit with a model describing competitive inhibition, only these fits areshown.

ing from �80 mM with potassium fluoride to �1 mM with potas-sium iodide. Similar results were obtained for the N. crassa NMO,with inhibition constants ranging from �125 mM for potassiumfluoride inhibition to �5 mM for potassium iodide inhibition. Nei-ther enzyme was inhibited by potassium phosphate at concentra-tions as high as 100 mM, suggesting that either dianions oranions with an ionic volume of 90 Å3 [27] are unable to bind atthe active site of both the NMOs tested.

A plot of the pKis values as a function of the ionic volume of theinorganic anions tested as inhibitors in the range from 25 to 64 Å3

(Fig. 2) yielded a straight line with both enzymes, with the excep-tion of sodium nitrite. The experimentally determined inhibitionconstants for sodium nitrite (Table 1) were between two- tothree-times lower than the values that can be predicted from thesize of the inorganic anion by lines of Fig. 2 of 4.7 mM for theH. mrakii and 19 mM for the N. crassa enzymes. Inhibition of theH. mrakii enzyme was more sensitive to the volume of the inor-ganic anion as indicated by the slope of 0.041 ± 0.001 mM/Å3 inthe plot of pKis versus ionic volume of the anion used as a compet-itive inhibitor for the enzyme, as compared to 0.027 ± 0.001 mM/Å3 for the N. crassa NMO (Fig. 2). The y-intercepts determined fromthe linear fitting of the data as shown in Fig. 2 were similar to oneanother irrespective of the enzyme used, with a value of ��2.8 mM. No correlation between the electron affinities of theinorganic anions tested and the inhibition constants of NMO wasfound, as illustrated in Fig. 2. The counterions of the inhibitorstested had no effect on the inhibition of either enzyme as indicatedby the inhibition constants of sodium nitrate, which conformed tothe linear relationship between the ionic volumes of the anionstested using potassium salts with the inhibition constants of theNMO.

3.3. Contributions of the alkyl chain length to substrate binding inNMO

In order to establish whether the hydrocarbon chain of the alkylnitronate substrate is a determinant for binding to the active site ofthe enzymes, the dissociation constants for a number of nitronatesubstrates and aliphatic aldehydes inhibitors with hydrocarbonchains of various lengths were determined in 50 mM potassiumphosphate at pH 7.4 and 30 �C. As shown in Fig. 3 for the case ofethylnitronate as substrate, stopped-flow measurements demon-strated that the reductive half reaction of the H. mrakii NMO in-volves the formation of an anionic flavosemiquinone as evidentfrom the peaks centered at �372 and 490 nm in the UV–visibleabsorbance spectrum obtained after anaerobic mixing of the en-zyme with the substrate. The observed rates of flavin reduction in-

Table 2Reductive half reaction of H. mrakii NMO at pH 7.4 at 30 �Ca.

Substrate kred (s�1) Kd (mM)

Ethylnitronate 210 ± 5 5.9 ± 0.3Propyl-1-nitronate 200 ± 5 5.3 ± 0.3Butyl-1-nitronate 350 ± 5 5.7 ± 0.2

a H. mrakii NMO was anaerobically mixed with substrate in 50 mM potassiumphosphate in a stopped-flow spectrophotometer and the absorbance changes at372 nm were monitored over time.

Table 3Aldehyde inhibition of H. mrakii NMO with respect to ethylnitronate as substratea.

Inhibitor Kis (mM)

Propanal 58 ± 3Butanal 60 ± 1Pentanal 58 ± 3Hexanal 60 ± 2

a Enzymatic activity was measured at varying concentrations of ethylnitronateand several fixed concentrations of inhibitor in 50 mM potassium phosphate pH 7.4at 30 �C. Data were fit with Eq. (3). The steady state kinetic parameters had thefollowing average values: kcat = 185 ± 3 s�1; Km = 3.3 ± 0.2 mM; kcat/Km = 55,000 ± 2200 M�1 s�1.

Fig. 2. Anion inhibition of NMO with respect to ethylnitronate as substrate. Theinhibition constants (Kis) for a series of monovalent anions were determined bymeasuring enzymatic activity of H. mrakii (�) or N. crassa (s) NMO at varyingconcentrations of ethylnitronate in the presence of several fixed concentrations ofinhibitor. All assays were carried out in 50 mM potassium phosphate at pH 7.4 and30 �C. Top panel: plots of pKis versus ionic volume. Values for the ionic volumes ofthe anions used were taken from [27]. Bottom panel: plots of pKis versus electronaffinity. Values for the electron affinity were taken from [31].

Fig. 3. Pre-steady state kinetics of H. mrakii NMO with ethylnitronate as substratein 50 mM potassium phosphate pH 7.4 and 30 �C. Flavin absorbance at 372 nm wasmonitored over time after anaerobic mixing of the enzyme with ethylnitronate togive final concentrations of 13 lM enzyme and 0.5 mM ethylnitronate. The UV–visible absorbance spectra before (black) and �30 s after mixing (red) are shown.Inset: Rates of flavin reduction determined from the fits of the stopped-flow tracesat 372 nm to Eq. (1) were plotted as function of ethylnitronate concentration (0.5 to50 mM) and were fit with Eq. (2). (For interpretation of the references to color inthis figure legend, the reader is referred to the web version of this paper.)

170 K. Francis, G. Gadda / Bioorganic Chemistry 37 (2009) 167–172

creased hyperbolically with the concentration of nitronate andshowed similar apparent Kd values of �5 mM with ethylnitronate,propyl-1-nitronate and butyl-1-nitronate, suggesting that thelength of the alkyl chain of the substrate is not a determinant factorfor binding of the substrate to the enzyme. The kinetic parametersfor the reductive half reaction of H. mrakii NMO are summarized inTable 2.

Each aldehyde tested was a competitive inhibitor with respectto ethylnitronate for the H. mrakii enzyme with inhibition con-stants of �60 mM, which further suggests that the alkyl chainlength of the substrate contributes only minimally with respectto the nitro moiety (Kis � 2 mM) for binding of the substrate tothe enzyme (Table 3). Similar, results were obtained with theN. crassa enzyme with propanal, butanal, pentanal or hexanal, inthat the inhibition constants for aldehyde inhibition were large(i.e., P70 mM), though accurate determinations could not be at-tained due to limited solubility of the inhibitors in aqueoussolution.

4. Discussion

Binding of alkyl nitronates to NMO is primarily mediatedthrough interactions of an anion binding pocket within the activesite of the enzyme and the nitro moiety of the substrate. Evidencesupporting this conclusion comes from inhibition studies of theH. mrakii and N. crassa enzymes with a series of monovalent, inor-ganic anions of varying ionic volumes. Binding of the anions likelyoccurs at the active site of the enzymes as evident from the spec-troscopic changes of the FMN cofactor induced upon incubationof the enzyme with nitrite, which resemble those induced by bind-ing of anthranilate to the active site of L-amino acid oxidase [28],and the competitive inhibition pattern observed with respect toethylnitronate as substrate. The linear correlation between the io-nic volumes of the inorganic anions used as inhibitors and the inhi-bition constants that describe their binding to the enzymessuggests the presence of a discrete binding pocket with a well-de-fined size, where the interactions at the anion binding site of thetwo enzymes between the ligand and the enzyme are maximizedas the volume of the inorganic anion increases. The lack of inhibi-tion of either enzyme by phosphate suggests that the anion bind-ing pocket can accommodate either monovalent ligands, but notdivalent ones, or only those with an ionic volume smaller than90 Å3 [27] which, by assuming a spherical geometry of the inor-ganic anion, corresponds to a diameter of the monovalent anionthat is smaller than 5.6 Å. In agreement with the estimate of thebinding pocket deriving from our kinetic analysis, the availablethree dimensional structure of NMO from P. aeruginosa clearlyindicates that an anion binding site is present at the active site of

K. Francis, G. Gadda / Bioorganic Chemistry 37 (2009) 167–172 171

the bacterial enzyme in proximity of the nitro moiety of the sub-strate [8], as illustrated in Fig. 4. In that enzyme, the anion bindingpocket is comprised mainly of the side chain of His152, the peptidylnitrogen of Gly151 and the side chain of Ser288. Although the crystalstructures of the enzymes studied in this report are not yet avail-able, the alignment of the amino acid sequences of the three en-zymes show that these residues are conserved in all three NMOs[1]. In the H. mrakii enzyme, the amino acid residues correspondingto the anion binding pocket of the bacterial enzyme are His197,Gly196 and Ser351. The equivalent residues in the N. crassa enzymeare His196, Gly195 and Ser342 (or Thr344).

A comparison of the slopes in the plots of pKis versus ionic vol-umes for the inorganic anions acting as inhibitors of the H. mrakiiand N. crassa NMOs is consistent with the anion binding pocket ofthe H. mrakii enzyme being slightly smaller than that of the N. cras-sa enzyme. In this respect, the slopes of the lines that fit the data inthe plots of Fig. 2 depend on either the values of the inhibition con-stants that describe the binding of the inorganic anions to the en-zyme, the ionic volumes of the inorganic anions, or both. Inprinciple, the binding of the inorganic anions to the binding pock-ets in the two enzymes can be affected by a number of factorsincluding geometric, steric, or electrostatic effects. However, theobservation that the fits of the data in Fig. 2 yields values for they-intercepts for the two enzymes that are similar to one anotheris consistent with the slope effect that is experimentally observedbeing due primarily to the volume of the anion binding pockets ofeach enzyme rather than other factors. The accurate determination

Fig. 4. Anion binding site in P. aeruginosa NMO. An electrostatic potential map wasgenerated for the X-ray crystallographic structure of P. aeruginosa NMO in complexwith 2-nitropropane (PDB ID: 2GJN) using an Adaptive Poisson–Boltzmann Solver[32] and visualized using Pymol. Panel A: Electrostatic potential surface of the activesite of P. aeruginosa NMO. Panel B: Active site amino acid residues that comprise theanion binding site in P. aeruginosa NMO.

of the sizes and geometries of the anion binding pockets at the ac-tive sites of the enzymes from H. mrakii and N. crassa will have toawait the elucidation of the X-ray crystallographic structures of thetwo enzymes, which is currently ongoing in collaboration withWeber’s group at Georgia State University.

Substrate recognition by NMO is predominantly determined bythe interactions occurring at the active site binding pocket of theenzyme with the nitro group of the alkyl nitronate substrate withminimal, if any, hydrophobic interactions of the enzyme with thealkyl chain of the substrate. Evidence supporting this conclusioncomes from the comparison of the Kis values for aldehyde inhibi-tion with respect to nitrite inhibition, which shows that the formerare at least 10-times smaller than the latter with both the enzymestested. Lack of interaction of the alkyl chain of the ligand with theenzyme is independently supported by the similar values for thedissociation constants for substrate binding (Kd) determined forthe H. mrakii enzyme with alkyl nitronates of varying chain lengthsof between two and four carbon atoms. Indeed, one would expect aprogressive decrease in the Kd values for the substrate withincreasing lengths of the alkyl chain of the substrate if hydrophobicinteractions played a significant role for substrate binding. The re-sults suggesting that NMO does not discriminate its ligands byexploiting the organic moiety of the substrate are consistent withprevious studies of the enzyme from N. crassa that established thatm-nitrobenzoate effectively binds to the enzyme (i.e., Kis value of9.1 mM at pH 7.4), despite its large size and aromatic character[4]. Further in agreement with minimal contribution of the alkylchain of the substrate to binding are previous studies of the H. mra-kii and N. crassa enzymes showing that the kcat/Km values for nitr-onates ranging from two to six carbon atoms are independent ofthe alkyl chain length of the substrate [2,3]. As illustrated inFig. 4, the three dimensional structure of the bacterial enzymefrom P. aeruginosa with 2-nitropropane bound at the active siteshows the presence of a wide cavity in the active site of the en-zyme, which is large enough to accommodate substrates of variouslengths or different structures [8].

5. Conclusions

In conclusion, the results presented herein demonstrate that theactive site of yeast NMO contains an anion binding pocket, whichparticipates in the binding of the nitro group of the alkyl nitronatesubstrates. Thus, the nitro group of the substrate is the key deter-minant for binding of the substrate at the active site of the enzymeas opposed to the hydrocarbon chain of the nitronate molecule act-ing as substrate, which plays a minimal role, if any, in binding bythe enzyme. These results contrast those previously reported fornitroalkane oxidase, whose ability to bind substrates at the activesite increases with increasing lengths of the alkyl chain of the sub-strate and reaches a maximum value with substrates containingfour or more carbon atoms [24]. A study of the pH and kinetic iso-tope effects on nitroalkane oxidase revealed that each methylenegroup of the substrate provides approximately 2.6 kcal/mol ofbinding energy [23], which was recently explained through struc-tural studies of the enzyme that demonstrated a hydrophobicchannel leading to the active site of the enzyme [12,29,30].Although structural studies have yet to be reported for the H. mra-kii and N. crassa enzymes, the X-ray crystallographic structure of P.aeruginosa NMO shows a solvent accessible active site that lacks ahydrophobic channel like that seen in nitroalkane oxidase [8]. Theelucidation of binding pockets for anionic ligands in the active siteof enzymes through inhibition studies with inorganic anions ofvarious ionic volumes demonstrates a kinetic method that shouldbe generally applicable to any enzyme whose crystallographicstructure is not yet available.

172 K. Francis, G. Gadda / Bioorganic Chemistry 37 (2009) 167–172

Acknowledgments

The authors thank Ms. Slavica Mijatovic and Ms. Nicole Chap-man for carrying out preliminary anion inhibition studies of theH. mrakii and N. crassa enzymes.

References

[1] G. Gadda, K. Francis, Arch. Biochem. Biophys. (2009), doi:10.1016/j.abb.2009.06.018.

[2] K. Francis, B. Russell, G. Gadda, J. Biol. Chem. 280 (2005) 5195–5204.[3] S. Mijatovic, G. Gadda, Arch. Biochem. Biophys. 473 (2008) 61–68.[4] K. Francis, G. Gadda, Biochemistry 45 (2006) 13889–13898.[5] K. Francis, G. Gadda, Biochemistry 47 (2008) 9136–9144.[6] K. Francis, G. Gadda, Biochemistry 48 (2009) 2403–2410.[7] M. Tchorzewski, T. Kurihara, N. Esaki, K. Soda, Eur. J. Biochem. 226 (1994) 841–

846.[8] J.Y. Ha, J.Y. Min, S.K. Lee, H.S. Kim, J. Kim do, K.H. Kim, H.H. Lee, H.K. Kim, H.J.

Yoon, S.W. Suh, J. Biol. Chem. 281 (2006) 18660–18667.[9] P.F. Fitzpatrick, A.M. Orville, A. Nagpal, M.P. Valley, Arch. Biochem. Biophys.

433 (2005) 157–165.[10] G. Gadda, P.F. Fitzpatrick, Biochemistry 39 (2000) 1406–1410.[11] G. Gadda, P.F. Fitzpatrick, Biochemistry 39 (2000) 1400–1405.[12] A. Nagpal, M.P. Valley, P.F. Fitzpatrick, A.M. Orville, Biochemistry 45 (2006)

1138–1150.[13] M.P. Valley, S.E. Tichy, P.F. Fitzpatrick, J. Am. Chem. Soc. 127 (2005) 2062–2066.[14] C.S. Hornfeldt, W.H. Rabe 3rd, J. Toxicol. Clin. Toxicol. 32 (1994) 321–324.[15] C. Kohl, A. Gescher, Xenobiotica 27 (1997) 843–852.

[16] C. Kohl, K. Mynett, J.E. Davies, A. Gescher, J.K. Chipman, Mutat. Res. 321 (1994)65–72.

[17] G. Shepherd, J. Grover, W. Klein-Schwartz, J. Toxicol. Clin. Toxicol. 36 (1998)613–616.

[18] E.S. Fiala, R. Czerniak, A. Castonguay, C.C. Conaway, A. Rivenson,Carcinogenesis 8 (1987) 1947–1949.

[19] P. Kreis, S. Brandner, M.W. Coughtrie, U. Pabel, W. Meinl, H. Glatt, U. Andrae,Carcinogenesis 21 (2000) 295–299.

[20] R. Ballini, G. Bosica, D. Fiorini, A. Palmieri, M. Petrini, Chem. Rev. 105 (2005)933–971.

[21] R. Ballini, A. Palmieri, L. Barboni, Chem. Commun. (Camb) (2008) 2975–2985.

[22] R. Ballini, M. Petrini, G. Rosini, Molecules 13 (2008) 319–330.[23] G. Gadda, D.Y. Choe, P.F. Fitzpatrick, Arch. Biochem. Biophys. 382 (2000) 138–

144.[24] G. Gadda, P.F. Fitzpatrick, Arch. Biochem. Biophys. 363 (1999) 309–313.[25] R.D. Allison, D.L. Purich, Method. Enzymol. 63 (1979) 3–22.[26] A.T. Nielsen, in: H. Feuer (Ed.), The Chemistry of the Nitro and Nitroso Groups,

vol. 1, Interscience Publishers, New York, 1969, pp. 349–486.[27] Y. Marcus, H.D.B. Jenkins, L. Glasser, J. Chem. Soc. Dalton (2002) 3795–3798.[28] P. Macheroux, O. Seth, C. Bollschweiler, M. Schwarz, M. Kurfurst, L.C. Au, S.

Ghisla, Eur. J. Biochem. 268 (2001) 1679–1686.[29] A. Heroux, D.M. Bozinovski, M.P. Valley, P.F. Fitzpatrick, A.M. Orville,

Biochemistry 48 (2009) 3407–3416.[30] A. Nagpal, M.P. Valley, P.F. Fitzpatrick, A.M. Orville, Acta Crystallogr. D Biol.

Crystallogr. 60 (2004) 1456–1460.[31] J.C. Rienstra-Kiracofe, G.S. Tschumper, H.F. Schaefer 3rd, S. Nandi, G.B. Ellison,

Chem. Rev. 102 (2002) 231–282.[32] N.A. Baker, D. Sept, S. Joseph, M.J. Holst, J.A. McCammon, Proc. Natl. Acad. Sci.

USA 98 (2001) 10037–10041.